Teaching Visual Orientation Discrimination Through Tactile Learning

Dingzhi Hu,Guozhen Liu,Lihan Chen,Cong Yu
DOI: https://doi.org/10.1167/18.10.91
2018-01-01
Journal of Vision
Abstract:Our previous double training studies demonstrate that perceptual learning can generalize to not only new locations and orientations (Xiao_et_al., 2008; Zhang_et_al., 2010), but also new physical stimuli encoded by different neural substrates (Wang_et_al., 2016). These results suggest that perceptual learning involves learning of sensory concepts (e.g., an orientation concept). Here we further show that such a concept can be represented and improved across sensory modalities. The participants first practiced tactile orientation discrimination of a 3D-printed grating at 40o or 130o with index fingers. Although training improved tactile orientation thresholds (35.2%, p=0.029), it didn't affect visual orientation discrimination at the trained orientation (p=0.15) and an orthogonal orientation (p=0.09). The participants then practiced an irrelevant visual contrast discrimination task at the trained orientation. After that, visual orientation discrimination was improved at both trained (33.4%, p< 0.001) and untrained orientations (22.9%, p=0.011). Further training of visual orientation discrimination failed to produce additional gain (-7.3%). In a second experiment, tactile orientation and visual contrast tasks were trained in the same sessions in alternating blocks of trials, which improved visual orientation discrimination at both trained (39.8%, p< 0.001) and untrained orientations (27.5%, p=0.004). Further training of visual orientation did not improve visual orientation thresholds (1.6%, p=0.80). These results indicate that tactile orientation learning can transfer completely to visual orientation with double training. In addition, we found that tactile orientation learning can transfer to an orthogonal tactile orientation with double training. These results provide further support for our claim that perceptual learning is concept learning (Wang_et_al., 2016), and suggest that stimulus orientation as a concept can be represented and improved through training at a modality-independent level. Learning transfers to a new sensory modality after the learned concept functionally connects to sensory inputs from the new modality through double training. Meeting abstract presented at VSS 2018
What problem does this paper attempt to address?